Table 3 Vehicle detection results of the VDTC-CEOADL algorithm in ISPRS Potsdam dataset.
Class | Accuy | Sensy | Specy | Fscore | AUCscore |
|---|---|---|---|---|---|
Training phase (80%) | |||||
 Car | 97.21 | 99.12 | 83.10 | 98.43 | 91.11 |
 Truck | 99.33 | 71.43 | 99.77 | 76.92 | 85.60 |
 Van | 98.33 | 89.19 | 99.15 | 89.80 | 94.17 |
 Pickup car | 99.00 | 54.05 | 99.94 | 68.97 | 77.00 |
 Average | 98.47 | 78.45 | 95.49 | 83.53 | 86.97 |
Testing phase (20%) | |||||
 Car | 98.22 | 99.02 | 90.24 | 99.02 | 94.63 |
 Truck | 99.55 | 60.00 | 100.00 | 75.00 | 80.00 |
 Van | 99.11 | 100.00 | 99.04 | 94.29 | 99.52 |
 Pickup car | 99.55 | 33.33 | 100.00 | 50.00 | 66.67 |
 Average | 99.11 | 73.09 | 97.32 | 79.58 | 85.20 |
